The adhesion of transformed cells, a property which potentially plays an important role in metastasis, is defective when compared with their normal counterparts. The basis for this defect is being sought in terms of changes in the cell-to-substrate contacts developed by transformed cells. The contacts are evaluated by interference reflection microscopy with emphasis on the stronger focal contact. The relative roles of different cell surface glycoconjugates in forming substrate contacts have been evaluated through reconstitution and inhibitor experiments on an adhesion-defective mutant and its wild-type. Collectively, the experiments show a specific requirement for glycosylated cell surface proteins, as opposed to gangliosides or proteoglycans, in the ability to form the focal contact. A functional role for substrate-adsorbed fibronectin, as opposed to other spreading factors such as platelet factor 4, has been demonstrated in the formation of the focal contact. Neither the cell-binding nor heparin-binding proteolytic fragments of FN, alone or in mixtures, possesses this property which must reside at some higher level of structural order in the intact dimer. This is under investigation. The results are significant for metastasis and the proteolytic activity of tumorigenic cells. Changes in a 92 kilodalton glycoprotein and 75 kilodalton protein that occur in the AD6 mutant of 3T3 cells which lack focal contacts and were originally thought to be characteristic of transformed cells have been induced in normal cells by glucose starvation without an effect on the focal contacts, thus eliminating a role of these proteins in adhesion. Proteins specifically linked to actin have been isolated from NP-40 solubilized fibroblasts. A 38 kilodalton protein is characteristic of cells with only focal contacts, and 150 kilodalton and 160 kilodalton proteins are characteristic of cells with only close contact. The glycosylation status and external exposure of the proteins are being examined, and similar differences are being sought in normal/transformed cell pairs. Preliminary studies on the development of the cytoplasmic precursor of the focal contact confirm its presence at the ultrastructural (HVEM) level in the unattached, extending cell margin and in the form of a bundle of parallel microfilaments. Defects in the various steps in forming the cytoplasmic precursor and in the required cell surface glycoconjugates will be sought in order to explain changes in the substrate contacts and hence adhesion of transformed cells. (A)